The multi-wavelength Raman lidar PollyXT has been
regularly operated aboard the research vessel
Three case studies provide an overview of the aerosol detected over the Atlantic Ocean. In the first case, marine conditions were observed near South Africa on the autumn cruise PS95. Values of optical properties (depolarisation ratios close to zero, lidar ratios of 23 sr at 355 and 532 nm) within the MBL indicate pure marine aerosol. A layer of dried marine aerosol, indicated by an increase of the particle depolarisation ratio to about 10 % at 355 nm (9 % at 532 nm) and thus confirming the non-sphericity of these particles, could be detected on top of the MBL. On the same cruise, an almost pure Saharan dust plume was observed near the Canary Islands, presented in the second case. The third case deals with several layers of Saharan dust partly mixed with biomass-burning smoke measured on PS98 near the Cabo Verde islands. While the MBL was partly mixed with dust in the pure Saharan dust case, an almost marine MBL was observed in the third case.
A statistical analysis showed latitudinal differences in the optical properties within the MBL, caused by the down-mixing of dust in the tropics and anthropogenic influences in the northern latitudes, whereas the optical properties of the MBL in the Southern Hemisphere correlate with typical marine values. The particle depolarisation ratio of dried marine layers ranged between 4 and 9 % at 532 nm.
Night measurements from PS95 and PS98 were used to illustrate the potential of aerosol classification using lidar ratio, particle depolarisation ratio at 355 and 532 nm, and Ångström exponent. Lidar ratio and particle depolarisation ratio have been found to be the main indicator for particle type, whereas the Ångström exponent is rather variable.
Aerosols, solid or liquid particles dispersed in air, play an important role
in the Earth's climate system. By scattering and absorbing solar and
terrestrial radiation, aerosols highly affect the radiation fluxes and thus
the radiative budget. In addition to this direct aerosol radiative forcing, aerosols
also modify the microphysical properties of clouds by acting as cloud
condensation or ice nuclei and thereby influence the radiative budget
indirectly
As the impact of aerosols on the climate system is various, it has
to be considered in climate modelling to receive accurate results,
which is, however, challenging because not all aerosol types
contribute to the aerosol radiative forcing in the same way
Light detection and ranging (lidar) represents a key method to
investigate vertically resolved aerosol properties. Measurements
with high spatial and temporal resolution and under ambient
conditions are possible up to an altitude of 100 km depending on
the lidar set-up
RV
The first cruise with PollyXT on the RV
The lidar measurements during the Atlantic cruises were performed with the
portable Raman and polarisation lidar system
PollyXT–OCEANET. A detailed description of the optical
set-up can be found in
The latest set-up of PollyXT–OCEANET enables the
measurement of backscatter coefficient profiles at 355, 532, and 1064 nm and
extinction coefficient profiles at 355 and 532 nm. Furthermore,
depolarisation measurements at 355 and 532 nm are possible. A second
detection unit enables measurements near the lidar at 355 and 532 nm and the
corresponding Raman wavelengths 387 and 607 nm down to about 120 m above
the lidar
The backscatter coefficient
The emitted laser light of the PollyXT lidar is linear polarised. In the atmosphere, the light is depolarised when scattered by non-spherical particles like dust or ice crystals. The detected light therefore contains a cross-polarised component in addition to the parallel-polarised light and can be detected separately. The ratio of cross-polarised to parallel-polarised light backscattered by particles is called particle depolarisation ratio. If the particles are mainly spherical, the particle depolarisation ratio is about zero because the linear polarised light has been returned to the lidar without changing the polarisation state. Non-spherical particles show higher depolarisation ratios. This quantity therefore enables the determination of the particle sphericity.
Ångström exponent, lidar ratio, and depolarisation ratio are indicators of the aerosol type. By knowing typical values of the lidar ratio, Ångström exponent, and particle depolarisation ratio, the dominant particle type can be specified.
The retrieval of those lidar-derived parameters from
PollyXT measurements and the corresponding error
estimation are described in detail by
For the data analysis in this study vertical smoothing lengths between 127
and 457 m were applied depending on the signal-to-noise ratio. Details are
given within the figures. GDAS1 (Global Data Assimilation System) data were used for the data analysis as
soundings upon RV
Observational overview of the autumn cruise PS95 from Bremerhaven to
Cape Town. Time series of the 500 nm daily mean AOT and daily mean
440–870 nm Ångström exponent derived with Microtops sun-photometer
measurements
The temporal development of the range-corrected signal (i.e. the uncalibrated
attenuated backscatter signal) of the autumn transit cruise PS95 is shown in
Fig.
Same as Fig.
The spring transit cruise PS98 started on 11 April 2016 in Punta
Arenas (Chile) and ended on 11 May 2016 in Bremerhaven. Time series
of the range-corrected signal and volume depolarisation ratio at
532 nm are shown in Fig.
Regular cruises across the Atlantic Ocean from north to south in the northern hemispheric autumn and from south to north in the northern hemispheric spring provided a large amount of lidar data over the Atlantic. Dust has been regularly observed in the northern tropics and subtropics west of the Saharan desert. The AOT at 500 nm, measured with a Microtops sun photometer, has been slightly higher in the Northern Hemisphere than in the Southern Hemisphere, which indicates a higher aerosol load in the former.
Marine conditions during PS95: time series of the Microtops
sun-photometer-derived AOT at 500 and 440/870 nm Ångström
exponent
Profiles averaged for 30 November 2015, 01:15–02:30 UTC. Backscatter coefficient and depolarisation ratios are smoothed with 127.5 m vertical length. Extinction coefficients, lidar ratios, and Ångström exponents are smoothed with 127.5 up to 242 m and afterwards with 367.5 m. Meteorological data from GDAS1 (30 November 2015, 00:00 UTC) and radio sounding measurements (29 November 2015, 12:00 UTC) are also presented.
Three night measurements from PS95 and PS98 were selected to present
typical atmospheric conditions by means of a detailed discussion of
the optical properties in the MBL and in lofted layers. First,
almost pure marine conditions with an overlying dried marine aerosol
layer during the autumn cruise PS95 are discussed. Second, a case
study on the same cruise but with Saharan dust near the Canary
Islands is presented. Third, a case during the spring cruise in 2016
(PS98) with Saharan dust and biomass-burning aerosol mixtures near
the Cabo Verde islands is shown. These three case studies are
marked with black stars on the cruise tracks
(Fig.
On 29 and 30 November 2015 at the end of the cruise PS95, clean conditions
could be observed near Cape Town. In this area, the dominant aerosol was of
marine origin according to CALIPSO aerosol classification
In Fig.
The time series of the volume depolarisation ratio (Fig. 4b) shows a thin layer of enhanced depolarisation at the top of the MBL at 300–400 m. This layer consists of dried marine particles and will be discussed later in this section.
Mean profiles of the measured optical properties are shown in
Fig.
NOAA HYSPLIT backward trajectories for 4 days ending at the
position of RV
The special highlight in this case study is the increase of the
depolarisation ratio at the top of the MBL, whereas the lidar ratio
within this layer is low, 16
Thus, we can conclude that marine particles were transported above the MBL top, dried, and crystallised and therefore cause a high particle depolarisation ratio even though the backscattering is low compared to the MBL.
This case confirms that marine aerosol can cause depolarisation in the lidar signal when RH is low. Without considering this property of marine aerosol, aerosol layers above the MBL causing depolarisation may be falsely classified. Automatic classification algorithms like the ones for CALIPSO, EarthCARE, and other lidars should take these feature into account, if relative humidity measurements are available, to not misclassify these aerosols as, for example, mixed dust.
First dust event observed during PS95: time series of the Microtops
sun-photometer-derived AOT (500 nm) and 440/870 nm Ångström
exponent
When RV
Figure
The range-corrected signal at 1064 nm and the 532 nm volume depolarisation
ratio of the first dust plume are shown in Fig.
Profiles averaged for 11 November 2015, 19:30–21:00 UTC. Dust
fraction calculated following
Averaged profiles of the measured optical properties and radio
sounding and GDAS1 profiles of temperature and relative humidity are
shown in Fig.
The MBL reached a height of about 400 m according to the backscatter
profile. Lidar ratios at 532 and 355 nm are 30
Complex aerosol layering with smoke and dust on PS98:
sun-photometer-derived AOT at 500 and 440/870 nm Ångström
exponent
Averaged profiles for 29 April 2016, 20:15–21:00 UTC. Dust and
smoke fractions calculated following
During the spring cruise PS98, extended aerosol layers with enhanced
depolarisation were observed near the Cabo Verde islands. The range-corrected
signal and volume depolarisation ratio at 532 nm as well as Microtops sun-photometer measurements on 29 April 2016 are shown in
Fig.
An increased backscatter coefficient at both wavelengths indicates
aerosol layers between 0.9 and 3 km. These layers are separated
from the MBL, which reached a height of about 500 m according to
the increased backscatter signal and the GDAS1 and radio sounding
data. The mean lidar ratio at 355 nm is 22
Mean profiles of the optical properties averaged from 29 April 2016
between 20:15 and 21:00 UTC are shown in
Fig.
The second layer extends from 1.3 to 1.6 km. The mean lidar ratio is
57
In the third layer, the mean lidar ratio is 40
Between 2.3 and 2.5 km, in the fourth layer, the particle
depolarisation ratio rises again (18
The fifth layer is characterised by a high lidar ratio up to 88 sr at
532 nm and 68 sr at 355 nm and high backscatter and extinction-related
Ångström exponents of 0.4
Figure
NOAA HYSPLIT backward trajectories ending at 29 April 2016
21:00 UTC at the position of RV
Mean values of extinction coefficient, lidar ratio and particle
depolarisation ratio at 532 nm, and the backscatter-related Ångström
exponent at 355/532 nm (top down) for MBL (blue), elevated aerosol
layers (black), and dried marine layers (red) on PS95
During this night measurement, five layers with different fractions of dust and smoke could be detected. At the same time, the MBL was almost pure marine without mixed-in dust or smoke particles. This case study shows that the MBL is not always influenced by dust and smoke transport and different aerosol types can occur at the same time above the Atlantic.
A statistical analysis of all Raman measurements with suitable weather conditions and signal quality was performed to provide an overview of latitudinal differences and characteristics of the different aerosol types observed over the Atlantic. A total of 45 night measurements from PS95 and PS98 were selected for analysis with respect to optical aerosol properties. Each measurement was screened for separated aerosol layers. The MBL and, when present, elevated aerosol layers and layers of dried marine aerosol, as presented in the first case study, have been analysed separately. These layers with enhanced depolarisation ratio directly above the MBL will be named dried marine layers.
The MBL top height and the extent of analysed elevated aerosol layers and
dried marine layers are shown in the first row of
Fig.
Mean MBL lidar ratios during PS95 are around 25
Elevated aerosol layers were mainly observed between 30
Latitudinal differences can also be seen in the course of the
backscatter-related Ångström exponent at 355/532 nm in the
MBL during PS95. The mean Å
The most prominent feature of dried marine layers is the enhanced
particle depolarisation ratio of about 4–9 % compared to the MBL with depolarisation ratios below 3 %. Those values are similar to
previous observations by
Differences in optical aerosol properties between northern and southern latitudes and the dust-influenced region west of the Saharan desert were detected. Whereas the Northern Hemisphere is influenced by anthropogenic pollution, southern latitudes are more likely to be influenced by marine aerosols only. Nevertheless, pure marine conditions, not influenced by aerosol originating from the continent, are rare and could only be observed at the end of PS95 near South Africa. Mostly, low-level clouds at the top of the MBL at the southern latitudes prohibited the lidar data analysis and thus the evaluation of more cases of pure marine conditions. In about 65 % of the cruise time in the Southern Hemisphere and 50 % in total during both cruises, clouds along the cruise track did not allow lidar data analysis.
Lidar ratio as a function of the particle depolarisation ratio at
355 nm
Mean values of optical properties of the MBL and elevated aerosol layers from
PS95 and PS98 (shown in Fig.
A clear separation of marine and dust-influenced MBL measurements
can be seen. Pure marine MBL measurements show lidar ratios between
13 and 40 sr and particle depolarisation ratios less than 2.5 %
at 355 and 532 nm, whereas the particle depolarisation ratio of
dust-influenced MBL measurements ranges between 5 and 20 %, caused
by a significant amount of non-spherical particles in the MBL. The
lidar ratio within these layers also shows a tendency to higher
values with
increasing particle depolarisation, caused by dust particles.
Elevated aerosol layers can be divided into layers with a high
particle depolarisation ratio (20–30 %) and a lidar ratio of
about 50–60 sr, layers with a lidar ratio between 30 and 75 sr
and a moderate particle depolarisation ratio (
Therefore, to complete the picture of particle-type-dependent
optical properties, the backscatter and extinction-related
Ångström exponents at 355/532 nm and the
backscatter-related Ångström exponent at 532/1064 nm are
considered for particle type separation in addition to the lidar and
depolarisation ratio in Fig.
Resulting from the preceding investigations, we consider the lidar ratio together with the particle depolarisation ratio as best indicators for particle classification above the Atlantic. A clear characteristic in terms of lidar ratio and Ångström exponent for the dried marine layers is not visible. Further observations of those layers are needed to get a comprehensive picture of dried marine aerosol properties.
The values presented above might be valuable information for new aerosol typing schemes needing knowledge from marine areas at the specific lidar wavelengths as, for example, for the upcoming EarthCARE mission. The operated lidar will measure at 355 nm but also requires information on the spectral behaviour of the optical aerosol properties to obtain radiation closure, which is one goal of this mission.
Three-dimensional illustration of the relation between lidar ratio, depolarisation ratio, and backscatter and extinction-related Ångström exponent. Blue dots represent MBL measurements, black dots elevated aerosol layers, and red dots dried marine layers. Coloured ellipses denote the different aerosol categories as in Fig. 14.
Multi-wavelength Raman polarisation lidar measurements from two ship-borne
cruises across the Atlantic Ocean (meridional direction) were analysed.
Pure marine, pure dust, and dust–smoke mixed conditions were observed. The
MBL was often mixed with dust near the equator and northern subtropics,
whereas in the outer tropics the marine influence dominated. One highlight
was the observation of dried marine aerosol at the top of the MBL, which was
relatively often observed during the cruises aboard RV
A statistical analysis showed latitudinal differences and the potential for aerosol classification of these cruises. Optical properties in the MBL were influenced by down-mixing of dust in the tropics and anthropogenic sources in the northern latitudes. In the Southern Hemisphere, optical properties of the MBL correlate with typical marine values. The mixing of dust in the MBL was low, confirmed by a continuous particle depolarisation ratio of less than 1 % in the MBL in the Southern Hemisphere. On both cruises, the MBL top never exceeded 900 m. Elevated aerosol layers were mainly observed in the Northern Hemisphere tropics and reached up to 4 km. Layers of dried non-spherical marine aerosol on top of the MBL could be observed only a few times, since, in about 65 % of the time in southern hemispheric mid-latitudes, low-level clouds prohibited the processing of the lidar data for aerosol properties.
All 45 night measurements from PS95 and PS98 were used to illustrate dependencies between lidar ratio, particle depolarisation ratio, and Ångström exponent for the different aerosol types. Lidar ratio and particle depolarisation ratio are the main indicators for the characterisation of the particle types observed over the Atlantic, whereas the Ångström exponent is not a good indicator for aerosol typing. Marine, dust, and smoke aerosols could be clearly identified with particle depolarisation and lidar ratio. But care must be taken when layers of dried marine aerosol occur at the top of the MBL, as the enhanced depolarisation ratio (4–9 %) could lead to wrong conclusions about the mixing state of the aerosol by inferring the presence of mineral dust. We therefore recommend considering the relative humidity and the vertical connection to the marine boundary layer when performing aerosol typing over the ocean, e.g. by space-borne lidars such as CALIOP or EarthCARE.
The values obtained increase the knowledge of the aerosol conditions in marine environments which make 70 % of the Earth's surface. Therefore, the presented results may also be a valuable contribution for the data analysis of satellite retrievals, which are the only instruments able to cover this large part of the Earth at the moment. The obtained data can also be used to validate and further improve model calculations, for example, by evaluation of the height of the different aerosol layers. Nevertheless, future studies are needed to expand the knowledge of dried marine aerosol, its drying processes, and interactions with aerosols above and within the MBL.
Meteorological data of all RV
SB and HB performed the data analysis and led the manuscript writing.
RE, SB, and MR realised the experimental set-up on board the RV
The authors declare that they have no conflict of interest.
The authors acknowledge support through ACTRIS under grant agreement no. 262254
and ACTRIS-2 under grant agreement no. 654109 from the European Union's
Horizon 2020 research and innovation programme. We thank the Alfred Wegener
Institute and the RV